Selective Depletion of Adult GFAP-Expressing Tanycytes Leads to Hypogonadotropic Hypogonadism in Males

Abstract

In adult mammals, neural stem cells are localized in three neurogenic regions, the subventricular zone of the lateral ventricle (SVZ), the subgranular zone of the dentate gyrus of the hippocampus (SGZ) and the hypothalamus. In the SVZ and the SGZ, neural stem/progenitor cells (NSPCs) express the glial fibrillary acidic protein (GFAP) and selective depletion of these NSPCs drastically decreases cell proliferation in vitro and in vivo. In the hypothalamus, GFAP is expressed by α-tanycytes, which are specialized radial glia-like cells in the wall of the third ventricle also recognized as NSPCs. To explore the role of these hypothalamic GFAP-positive tanycytes, we used transgenic mice expressing herpes simplex virus thymidine kinase (HSV-Tk) under the control of the mouse Gfap promoter and a 4-week intracerebroventricular infusion of the antiviral agent ganciclovir (GCV) which kills dividing cells expressing Tk. While GCV significantly reduced the number and growth of hypothalamus-derived neurospheres from adult transgenic mice in vitro, it causes hypogonadotropic hypogonadism in vivo. The selective death of dividing tanycytes expressing GFAP indeed results in a marked decrease in testosterone levels and testicular weight, as well as vacuolization of the seminiferous tubules and loss of spermatogenesis. Additionally, GCV-treated GFAP-Tk mice show impaired sexual behavior, but no alteration in food intake or body weight. Our results also show that the selective depletion of GFAP-expressing tanycytes leads to a sharp decrease in the number of gonadotropin-releasing hormone (GnRH)-immunoreactive neurons and a blunted LH secretion. Overall, our data show that GFAP-expressing tanycytes play a central role in the regulation of male reproductive function.

Keywords: GnRH; adult neural stem/progenitor cells; glial fibrillary acidic protein; hypothalamus; reproduction; sexual behavior; tanycytes.

Figure 1

Distribution of GFAP-positive/Tk-positive cells in the MBH of GFAP-Tk male mice. Rostral (A), middle (B) and caudal (C) MBH showing for each panel, from left to right and up to down, confocal images of nuclei coloration with Hoechst (blue), Tk (green), GFAP (red), Hoechst/Tk, Hoechst/GFAP expression and the merge image. Insets are high magnification showing Tk (green), and GFAP (red) immunopositive cells. All the Tk-positive cells are located in the ependymal layer of the VMH and the dorsal AN, corresponding to the localization of α1 and a subset of α2 tanycyte subtypes. GFAP-positive/Tk-positive cells are also detected in the AN and ME parenchyma. VMH, ventro-medial hypothalamus; AN, arcuate nucleus; ME, median eminence; Tk, thymidine kinase. Scale bars, 20 µm; Insets: 5 µm.

Figure 2

GFAP-expressing cells are the predominant source of proliferative cells within the hypothalamus. Representative images of floating hypothalamic tertiary neurospheres (NS) derived from wild type mice (WT ctr; A), wild type mice receiving a ganciclovir treatment (WT+GCV; B), GFAP-Tk Tg mice (Tg ctr; C) and GFAP-Tk Tg mice treated with an i.c.v. ganciclovir injection (Tg+ GCV; D). Scale bar, 100 µm. Number (E) and size (F) of hypothalamic neurospheres after a 2-week culture period in presence of GCV. Data are expressed as the mean ± SEM, WT n=2 and Tg mice n>3 for each group, n=3 experiments. ***p<0.001.

Figure 3

In vivo central GCV administration in Tg mice impaired expression of NSPC markers within the MBH. Representative images of GFAP expression measured in the mediobasal hypothalamus of control (A) and Tg+GCV male mice (B). Statistical analysis revealed no variation of GFAP expression in the VMH, AN and ME (C). In contrast, the number of GFAP-positive tanycytes located in the ventricular border is statistically lower in Tg+GCV male mice as compared to control mice (D). Confocal images of Vimentin expression in the mediobasal hypothalamus of control (E) and Tg+GCV male mice (F). Tg+GCV male mice display a significantly decreased Vimentin expression in the VMH, NA and ME (G). Representative images of Sox2 expression in the mediobasal hypothalamus of control (H) and Tg+GCV male mice (I). Statistical analysis revealed significantly less Sox2 positive cells in the ventromedial hypothalamus (VMH) (J), in the AN and in the ME (K) in Tg+GCV male mice. Data are expressed as the mean ± SEM, n≥5 for each group. *p<0.05, **p<0.01. ***p<0.001, ^#p=0.06 or 0.07. Scale bars, 50 µm.

Figure 4

Alteration of gonadal functions after central GCV administration in Tg mice. Tg+GCV male mice display decreased testicular weight (A) and plasma testosterone concentrations after hCG (human chorionic gonadotropin) injection (B). Testicular histology of WT ctr (C), WT+GCV (D), Tg ctr (E) and Tg+GCV (F) male mice revealed alterations and vacuolization of seminiferous tubules after central GCV administration in Tg mice, scale bars, 50 µm Insets show a higher magnification image of a single seminiferous tube for each group of mice. Scale bar, 200 µm. Tg+GCV male mice display decreased serum LH concentration (G) while serum FSH concentration is not affected by the treatment (H). Data are expressed as the mean ± SEM, n≥3 for each group. *p<0.05, **p<0.01, ***p<0.001.

Figure 5

Effect of central GCV administration in Tg mice on GnRH immunoreactivity in the preoptic region and the median eminence. (A, B) Representative images of GnRH fibers in the median eminence of control (A) and Tg+GCV (B) male mice. Statistical analysis revealed a significantly decreased GnRH expression in Tg+GCV mice as compared to the three control groups (C). (D, E) Representative images of GnRH neurons in the preoptic area (POA) of control (D) and Tg+GCV (E) male mice. (F) Mean number of GnRH neurons in the POA per slice, n≥3 slices per group. Statistical analysis revealed a significantly decreased GnRH neuron number in the POA in Tg+GCV male mice (F). Nuclei coloration with Hoechst (blue). Data are expressed as the mean ± SEM. *p<0.05. Scale bars, 20 µm.

Figure 6

Central GCV administration in Tg mice affected male sexual behavior. The latencies to first mount (A) and to first intromission (B) are increased in Tg+GCV male mice, while the intromission frequencies (C) are decreased after central GCV administration in Tg mice. Males of the four groups spend more time in close contact to a receptive female than an unfamiliar male (D). Data are expressed as the mean ± SEM, n≥6 for each group. *p<0.05, **p<0.01.